Human activities are a major driver of biodiversity degradation and loss, especially in tropical forest areas, where forest-fringe towns and villages depend on the forests for their livelihoods. In order to reduce threats that human activities pose to biodiversity, livelihoods support programs are employed as economic incentives for biodiversity conservation. These programs support the livelihoods activities of local communities, with the aim of triggering favorable attitudes and behaviors towards conservation, and ultimately reduce biodiversity degradation. Their effectiveness as conservation tools has not been evaluated. I investigated the effects of livelihoods programs on conservation attitudes and the consequent effects on biodiversity in the Afadjato-Agumatsa and Atewa forest areas in southeastern Ghana. The study areas are coupled human and natural systems, which are excellent for research in the theoretical framework of biocomplexity in the environment. Using literature reviews and field visits, I documented the specific livelihoods support activities (LSAs) used for biodiversity conservation, their historical trend and geographical distribution in Ghana. I used ex-post costbenefit analysis to determine socio-economic estimates of the LSAs in the two forest areas. Since communities were not randomly assigned to the interventions, I employed quasi-experimental design to evaluate the effects of LSAs on environmental attitudes. I evaluated the effect of conservation attitudes on biodiversity at two levels. These levels included 1) functional biodiversity at the landscape level represented by mean Normalized Difference Vegetation Index (NDVI) of forest; and 2) compositional biodiversity at the species level represented by species diversity of fruit bats. iv The earliest record of LSAs used for biodiversity conservation in Ghana was in 1993. I identified 71 different activities belonging to eight categories. Some of these activities are beekeeping, animal husbandry, crop farming, and snail rearing. Most LSA programs have been in northern Ghana. There was an increasing tendency to make LSAs part of every conservation program in Ghana and this satisfies the current policy of collaborative conservation. The socio-economic estimates of LSAs included: 1) capital investment; 2) net socio-economic benefits; and 3) the benefit-cost ratio. The per-community values of the three estimates were not different between the two study areas. The per capita values of capital investment and net economic benefit were not significantly different between the two study areas. However, benefitcost ratio per capita was higher in Afadjato-Agumatsa than in Atewa. Estimates of economic returns from LSAs were marginal but the perceptions of success were relatively high. Environmental attitudes in LSA communities and non-LSA communities were not significantly different, and this was confirmed by an estimate of infinitesimal effects of LSAs on forest conservation attitudes. Among LSA communities, benefit-cost ratio of LSAs predicted favorable forest conservation attitudes; and change in pro-conservation attitudes were significantly higher in communities that had active LSAs than in communities which had no active LSA. Mean NDVI of the forests decreased from 1991 to 2000 and decreased further but at a slower rate to 2010. Higher forest conservation attitudes predicted higher mean NDVI in 2010. Higher change in mean NDVI from 1991 to 2000 predicted higher change in mean NDVI from 2000 to 2010. Eleven of the 13 fruit bat species in Ghana were recorded in the study areas. Longer v distances between a local community and its forest predicted higher species diversity of forestspecialist fruit bats. The results indicate that LSAs have become a major contribution to Ghana’s current collaborative forest policy. The fact that perceptions of LSA success were moderate even though the economic returns from them were marginal suggest that other factors such as provision of employment, training in new skills and community cohesion played a part in how communities viewed the success as LSAs. Evaluations of conservation attitudes suggest that just participating in LSAs did not improve attitudes; but higher benefit-cost ratio predicted favorable conservation attitudes, and conservation attitudes were higher in communities that sustained their LSAs. Therefore, it may serve biodiversity conservation to invest in LSAs that can be sustained and involve the least costs to local communities. Primary production of the forests, a proxy for a functional habitat, continued to decrease. Preventing communities from locating closer to forests could improve fruit bat diversity, which contributes to natural forest regeneration. Improving conservation attitudes should be an objective of conservation at the landscape scale. On the basis of the results, I developed a conceptual model for forest biodiversity conservation in a biocomplexity framework. This model could be useful for evaluating conservation in tropical forest areas. Lessons from this study can be applied in other incentive-based conservation programs such as payments for ecosystem services systems and carbon market schemes. I suggest that this study be repeated after a decade and that other socio-political and biogeochemical variables be integrated into future studies.